Performance of Three Major Techniques for Endoscopic Submucosal Dissection: A Systematic Review and Network Meta-analysis

Introduction: Endoscopic submucosal dissection (ESD) has been recognized as the standard treatment for early malignant lesions in the gastrointestinal (GI) tract. limited evidence is synthesized on effectiveness different techniques employed to facilitate ESD. We assessed the comparative efficacy of ESD techniques through a network meta-analysis. Methods: Randomized controlled trials (RCTs) comparing different techniques for ESD, including tunnel/pocket method (Tu-ESD), traction method (Tr-ESD), and conventional method (C-ESD) were identified. Study outcomes were en-bloc resection, curative resection, procedure time, and adverse events. We performed network meta-analyses for all treatments and used Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) criteria to appraise the quality of evidence. Results: A total of 18 RCTs involving 2,677 patients were analyzed to evaluate the effectiveness of three major ESD techniques. Network meta-analysis results showed no inconsistencies across the network for all outcomes. According to SUCRA analysis, Tu-ESD achieved the highest ranking for curative resection (SUCRA score 93.7), whereas Tr-ESD ranked highest ranking for reducing procedural time (SUCRA score 98.7). Tr-ESD demonstrated a significant reduction in procedural time compared to C-ESD (risk difference: -19.15; 95% CI: -28.56 to -9.73). Sensitivity and subgroup analyses (according to colorectal, gastric, and esophageal locations) showed that Tr-ESD was best for en-bloc resection and procedure time, while Tu-ESD was best for curative resection as well as adverse events. Conclusion: Both Tu-ESD and Tr-ESD are effective and safe dissection methods compared to C-ESD. Given that different ESD techniques offer certain advantages, the choice of technique should be tailored to the specific clinical scenario.

Publikationsverlauf

Eingereicht: 24. April 2025

Angenommen nach Revision: 25. Juni 2025

Accepted Manuscript online:
25. Juni 2025